There’s a document circulating in the upper echelons of American power that deserves more serious philosophical attention than it has so far received. Palantir’s 22-point manifesto, drawn from Alex Karp’s The Technological Republic, has been described by scholars of authoritarianism as one of the most alarming public statements to emerge from Silicon Valley in recent memory. Cas Mudde called it “one of the scariest things I have seen in a while.” Speaking on the BBC earlier this week I called it “the most honest statement of technofascism we’ve yet seen in public”. Donald Moynihan identified its organising logic with precision: a US government and its Silicon Valley allies operating without accountability, unconstrained by democratic oversight, with Palantir’s financial interests perfectly aligned with a world that chooses military aggression over diplomacy. Yanis Varoufakis pushed to the ethical core: a company that regards ethics as a liability, equipping military forces with autonomous targeting systems designed to remove whatever residual moral judgment combatants might exercise on the battlefield.
These responses are correct as far as they go. But forensic alarm, however justified, is insufficient. What the manifesto requires is not more denunciation but deeper understanding — of where this thinking actually comes from, why it is more structurally dangerous than anything classical authoritarianism produced, and what a philosophically serious counter-claim looks like. Those are the three things this essay attempts.
The Doctrine and Its Ancestors
Peter Thiel has never hidden his intellectual debts. He has spoken openly about René Girard, whose mimetic theory he encountered at Stanford and has described as the most important idea he ever encountered. He has funded institutions associated with the Straussian tradition. And the fingerprints of Carl Schmitt — the German jurist who became the most sophisticated legal theorist of the Nazi state — are visible throughout the manifesto’s architecture, even if the name never appears.
These are not incidental influences. They form a coherent philosophical inheritance, and understanding it changes what the manifesto actually is.
Schmitt’s central contribution to political thought was the concept of the exception — the moment when normal legal and constitutional order is suspended in the face of existential threat. For Schmitt, sovereignty is not defined by who governs in ordinary times but by who decides when ordinary times have ended. The sovereign is whoever has the authority to declare the emergency. What the Palantir manifesto does, in effect, is declare that the emergency is permanent — that the existential threat posed by technological competition with China, by civilisational “dysfunction,” by the weakness of democratic institutions, is not a temporary condition requiring temporary measures but the constitutive condition of the present age. Permanent emergency is the justification for permanent exception. And permanent exception is the justification for permanent unaccountable power.
Strauss adds a different layer. His reading of classical philosophy — contested, but enormously influential among a particular strain of American conservative thought — held that genuine political wisdom cannot be made fully public. Democratic populations are not equipped to receive certain truths about power, necessity, and the fragility of civilisation. The philosopher-guardians who understand these truths have both the right and the obligation to govern in ways that the masses cannot fully comprehend or sanction. The manifesto’s contempt for pluralism, its civilisational rankings, its implication that some peoples are equipped for self-governance and others are not — this is Straussian condescension given a Silicon Valley inflection. The tech founders who understand the algorithmic future are the new guardians. The rest of us are the population to be managed for our own good.
Girard’s contribution is stranger and in some ways more revealing. His mimetic theory holds that human desire is fundamentally imitative — we want what others want because they want it — and that this mimetic rivalry inevitably generates violence, which societies contain through scapegoating mechanisms: channelling collective aggression onto a sacrificial victim. Thiel’s reading of Girard is idiosyncratic but consistent: he takes from it a profound suspicion of democratic consensus (which he reads as mimetic conformity) and a conviction that civilisational collapse is the default condition that only exceptional individuals — those who escape the mimetic trap — can arrest. This is not a neutral intellectual position. It’s a justification for the refusal of democratic constraint dressed in the language of anthropological insight.
Together, these three thinkers provide the manifesto with something more dangerous than political ambition: a philosophical system. The emergency is permanent. The guardians must govern without full democratic accountability. And those who resist are either mimetic conformists or civilisational liabilities. The doctrine is not improvised. It has foundations. And foundations are harder to dislodge than positions.
The End of the Need for Consent
Classical fascism was, among many other things, a performance. It required stadiums, torchlight, the choreographed merger of individual identity into collective will. Mussolini needed crowds. Hitler needed rallies. The emotional technology of twentieth-century authoritarianism was mass mobilisation — the manufacturing of a people who felt the programme as their own, who identified with it viscerally enough to enforce it on each other. This required, paradoxically, a kind of consent — not the procedural consent of liberal democracy, but the affective consent of a population that had been successfully persuaded to want what the regime wanted.
What is being built now requires none of this.The architecture Palantir is constructing operates below the threshold of popular consciousness. It doesn’t need the population to identify with the programme. It doesn’t need rallies or rhetoric or the careful manufacture of collective emotion. It needs data infrastructure, predictive modelling, algorithmic targeting, and the quiet integration of surveillance capacity into the administrative systems that populations already depend on and cannot easily exit. ICE doesn’t need a cheering crowd to locate and detain an undocumented person. It needs a database and a pattern-matching system. The IDF doesn’t need soldiers who have made a moral commitment to a particular target. It needs a targeting algorithm that has already made the decision before the soldier’s finger reaches the trigger.
This is not a superficial difference from classical fascism. It is a structural one. The great democratic achievement of the twentieth century — imperfect, contested, and perpetually under threat — was the recognition that political legitimacy requires popular consent, and that consent requires the possibility of refusal. The entire architecture of civil liberties, democratic procedure, and the rule of law exists to protect and institutionalise that possibility. You can say no. You can organise. You can vote the programme out.
What algorithmic governance does, at its limit, is make the possibility of refusal operationally irrelevant. Not illegal — irrelevant. The question is not whether the population approves of being surveilled but whether the population retains the practical capacity to do anything about it. When the surveillance is infrastructural — implanted in the systems through which you access healthcare, cross borders, use financial services, move through cities — the liberal distinction between consent and coercion begins to dissolve. Clearly, you haven’t consented. But you also haven’t been coerced in any way that existing legal frameworks can easily name or address.
Scholars of authoritarianism have a name for this: preference falsification — the gap between what people privately believe and what they publicly express under conditions of surveillance and anticipated reprisal. But what’s emerging now goes further. It’s closer to preference irrelevance. The system doesn’t need to know what you prefer, and it doesn’t particularly need to change what you prefer. It needs to process you — to sort, filter, flag, detain, or clear — regardless of what you think about it.
This is why the technofascism framing, while useful as provocation, requires challenge. Fascism, in its classical form, was a politics of passionate identification. What Palantir’s doctrine describes is a politics of passionate indifference to identification — a governance model that has technically evolved beyond the need for the theatre of consent. The danger is not that populations will be whipped into fervour. It’s that fervour will become unnecessary. That the machinery will simply run.
And here is where the structural logic becomes essential to understand. It’s one of the issues I pointed to in my book Teaching Silicon How to Feel. No individual within the system needs to intend this outcome. The engineers building the targeting systems are solving technical problems. The procurement officers authorising the contracts are managing budgets. The politicians approving the programmes are responding to security briefings. Each actor, operating within their own institutional logic, contributes to a system whose aggregate effect none of them chose and few of them have fully contemplated. The horror is not conspiracy. It is the banal, incremental, perfectly rational construction of something that has no rational exit once it reaches sufficient scale.
Sufficient scale may already have been reached.
What Technology Is For
Every doctrine of domination contains, implicitly, a theory of what human beings are. Palantir’s theory is not difficult to locate. It runs beneath the manifesto’s surface argument like an ice-cold current: human beings are mimetic, rivalrous, and organised into civilisational hierarchies of competence and dysfunction. The strong must govern the weak. The technologically capable must manage the technologically innocent. Order is imposed, not emergent. And the primary relationship between a state and its population — or between a dominant civilisation and the world it administers — is one of surveillance, sorting, and calibrated force.
This is not a new theory. It’s one of the oldest. Clannish indeed. What’s new is the infrastructure available to act on it at planetary scale.
A counter-doctrine must therefore begin not with policy but with anthropology. It must offer a different account of what human beings are, what they are capable of, and what technology is therefore for. Not a naively optimistic account — the evidence for human rivalry and civilisational dysfunction is not difficult to accumulate — but one that is more complete, more honest about complexity, and more adequate to the actual demands of the moment.
Here is the counter-claim: human beings are not primarily rivalrous. They are primarily relational. The deepest structures of human cognition, emotion, and social organisation are not competitive but cooperative — not because it’s morally preferable to competition, though it often is, but because it phylogenetically precedes competition. We evolved as creatures of radical interdependence. The capacity for language, for collective memory, and for the transmission of culture across generations are not achievements of exceptional individuals who escaped the mimetic trap. They are achievements of communities, of the slow accumulation of shared practice, of what the philosopher Merlin Donald called the external symbolic storage that makes civilisation possible at all. Thiel’s Girardian individualism mistakes the exception for the rule and the pathology for the norm.
From this different anthropology flows a different account of what technology is for. If human beings are primarily relational — if our deepest capacities are collective rather than individual, cooperative rather than rivalrous, syntrophic rather than extractive — then the measure of any technology is not its capacity to project force or process populations. It’s whether it deepens or degrades the relational fabric through which human life becomes meaningful and sustainable.
This is not a soft claim. It has hard edges.
A technology that embeds surveillance into the infrastructure of daily life doesn’t only threaten civil liberties in the abstract sense. It attacks the relational fabric directly — it introduces the permanent possibility of observation into spaces that require trust, spontaneity, and the freedom to be unfinished in order to function at all. Intimate life, creative practice, political deliberation, the formation of conscience — none of these can survive the condition of being permanently watched and processed. Surveillance doesn’t just monitor a population. It changes what that population can become.
A targeting algorithm that removes the combatant’s moral judgment from the act of killing doesn’t just raise questions about accountability under international law. It severs the last thread connecting violence to human conscience — the thread that has always, however imperfectly, carried the possibility of refusal, of mercy, of the soldier who does not pull the trigger. Once that thread is cut, it’s not simply that atrocities become easier to commit. It’s that the category of atrocity begins to dissolve, because atrocity requires a perpetrator capable of moral recognition, and the system has been designed to route around that capacity.
And a civilisational doctrine that ranks cultures as functional or dysfunctional, that regards some peoples as equipped for self-governance and others as requiring administration, doesn’t merely reproduce colonial logic. It forecloses the future — it mistakes the current distribution of technological and economic power for a permanent hierarchy of human worth, and in doing so, denies the possibility of the very civilisational transformation that the actual crises of this century require.
Because here is what the manifesto cannot accommodate, and what any serious counter-doctrine must place at its centre: we are living through a planetary emergency that no single civilisation, no dominant power, no algorithmic governance system can navigate alone. The ecological unravelling now underway does not respect civilisational hierarchies. The destabilisation of climate systems, the collapse of biodiversity, the exhaustion of the living substrate on which all human activity depends — these are not problems that American hard power, however AI-augmented, can solve by achieving dominance over its rivals. They are problems that require the relational capacities that Palantir’s doctrine systematically degrades: the capacity for genuine cooperation across difference, for the long-term thinking that industrial economism structurally prevents, for the recognition that future generations have claims on present decisions that no targeting algorithm will ever be programmed to honour.
The three obligations that constitute a genuinely different civilisational programme are not abstract ideals. They are functional requirements for navigating what is actually coming. Being genuinely beneficial to each person — not processing them, not sorting them, not administering them, but recognising their irreducible particularity and capacity. Being syntrophic for the living world — not extracting from it until the extraction becomes impossible, but participating in the regenerative cycles that make all life, including human life, sustainable. And honouring the claims of future generations — holding the decisions of the present accountable not only to current power distributions but to the world those decisions will leave behind.
This is what technology is for. Not the projection of dominance. Not the bypassing of consent. Not the management of populations ranked by civilisational worth. But the deepening of the relational fabric — the extension of human capacity to recognise, respond to, and take responsibility for each other and for the living world we share and have so far failed to tend.
The Summons
The Palantir manifesto names its programme openly. That is, in a grim way, a form of honesty — and it deserves to be met with equal honesty in return.
The counter-doctrine sketched here is not a policy platform. It will not be adopted by a procurement committee or written into a defence contract. What it is, is a different account of what human civilisation is for — one that takes the actual crises of this century seriously enough to refuse the consolations of dominance, and seriously enough to insist that the relational, cooperative, syntrophic capacities of human beings are not liabilities to be managed but the only genuine resources we have for what’s coming.
A doctrine of dominion only goes unanswered when there is nothing worth defending. I believe there’s much to defend. The manifesto has named its programme. The only adequate response is for us to name ours.